LLM EU
Menu
Mudelli

Gemma 4 12B It

Google DeepMind

Katalgu biss Piżijiet miftuħa

Aħħar reviżjoni:

Sovranità u ospitar

Katalgu biss

Dan il-mudell huwa kkatalgat għal referenza. Għadu mhux ospitat fuq LLM EU, u l-ebda endpoint tal-ewwel parti ma jservih hawn. It-talbiet lejh imorru direttament għand il-fornitur, taħt it-termini tal-fornitur stess.

Sommarju

Il-checkpoint huwa sintonizzat bl-istruzzjonijiet u jinkludi mudell ta' chat. Gemma 4 12B It huwa ppubblikat minn Google DeepMind bħala piżijiet miftuħa b'12B parametri taħt Apache-2.0. Jieħu awdjo u jikteb test. Il-konfigurazzjoni ppubblikata ma tirreġistra l-ebda tieqa ta' kuntest, għalhekk din il-karta ma tinkludi l-ebda waħda. Ir-repożitorju nħoloq f'Mejju 2026 u huwa rifless fuq Hugging Face. Il-karta hija biss għall-katalgu: LLM EU ma jħaddimx dan il-mudell. Il-pubblikatur ma jiddikjara l-ebda kopertura lingwistika fuq il-paġna tal-mudell.

Speċifikazzjonijiet

Fornitur
Google DeepMind (US)
Tieqa tal-kuntest
Mhux verifikat: il-fornitur ma ddokumentax din id-dikjarazzjoni.
Parametri
12B
Modalitajiet
text, vision, audio
Liċenzja
apache-2.0 · licence
Piżijiet miftuħa
yes
Data tar-rilaxx
23 Mej 2026
Kompiti
Appoġġ multilingwi, Viżjoni, Sommarizzazzjoni

Lingwi

Qawwi: —

Adekwat: —

Qawwi

    Adekwat

      Mhux verifikat: il-fornitur ma ddokumentax din id-dikjarazzjoni. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.

      Meta tużah

      Saħħiet

      • open weights are published on Hugging Face and can be downloaded without a vendor agreement
      • Apache-2.0 licence, so commercial self-hosting is permitted without a separate agreement
      • at 12B parameters it fits on a single accelerator
      • instruction-tuned checkpoint with a chat template in the repository
      • speech recognition runs locally, so audio never leaves the operator premises

      Limiti

      • catalogue only — LLM EU does not run this model, and nothing on this card implies an LLM EU endpoint
      • no evaluation has been run by LLM EU
      • the published configuration records no context window, so the card carries none
      • at 12B parameters, self-hosting needs a single accelerator with enough memory for the weights
      • vision checkpoints are heavier to serve than their parameter count alone suggests, because the image encoder and pre-processing run alongside the language model